A STUDY OF CERTAIN ASPECTS OF BLOODCOAGULATIONINTHE POSTOPERATIVESTATE IN CONGESTIVEHEART

FAILURE AND IN THROMBOPHLEBITIS

By NOBLE0. FOWLER,' WITH THE TECHNICAL ASSISTANCE OF SHIRLEY ROEHMAND HAROLDPERLMAN

(From the Cardiac Laboratory, Cincinnati General Hospital, and the Department of InternalMedicine, University of Cincinnati)

(Received for publication January 18, 1949)

In recent years a number of tests have been de-signed for the purpose of detection of acceleratedblood coagulation. Among these are the lusteroidtube coagulation test (1-4), the silicone tube co-agulation test (5), the modified clot retraction test(6), the modified Lee-White clotting test, employ-ing three test tubes instead of one (7), the heparinretarded coagulation test of Waugh and Ruddick(8, 9), the heparin tolerance test ( 10, 11 ), and theprothrombin time determination, using both wholeand dilute plasma (12-14).

The study reported herewith was undertakenfor the following purposes. First, it was desiredto ascertain whether or not there is any correlationamong the various tests themselves. Second, it wasdecided to determine whether any of these testscan be used to predict thrombophlebitis in patients.Finally, the incidence and duration of hypercoagu-lability in the postoperative state was studied, aswell as the incidence of hypercoagulability in con-gestive heart failure. Incident to doing these stud-ies, observations were made on the effect of vari-ation of tube size and temperature upon the clot-ting time of blood in the Lee-White test. Further,in doing the heparin tolerance test, blood waswithdrawn almost simultaneously from the veininto which the heparin was injected and from avein of the opposite arm. Clotting times were runon each sample and the results were compared.

MATERIAL

For the study on postoperative patients, a group of 29patients in the Cincinnati General Hospital who had hadrecent major surgical operations was selected at random.These 29 patients and three additional patients havingacute thrombophlebitis were used to determine the corre-lation among the various tests. The problem of hyper-coagulability in congestive heart failure was studied in 27

1 National Institute of Health Postdoctorate ResearchFellow.

patients suffering from cardiac insufficiency on the medi-cal wards of the Cincinnati General Hospital. A groupof house officers, nurses, technicians, and ambulatorypsychiatric patients were used as normal controls. Per-sons from both the control group and from the post-operative group were used to study the effect of variationin temperature and tube size upon the Lee-White clottingtest. Persons from both groups were also used to studythe effect of the veins chosen in performing the heparintolerance test.

METHODS

The 29 postoperative patients and the three patientshaving acute thrombophlebitis were each studied by meansof six tests which were run simultaneously. The sameseries of six tests was repeated on seven of the post-operative group later in their postoperative courses. Thesix tests employed were the modified Lee-White clottingtest, the lusteroid tube coagulation test, the prothrombintime on whole and on 12.5%o plasma, the Waugh-Ruddickheparin retarded coagulation test, and the heparin toler-ance test. The 27 patients suffering from congestiveheart failure were studied by means of the lusteroid tubecoagulation test and the modified Lee-White test usingthree dry test tubes and an oiled syringe. Fifty of thecontrol group were studied with the lusteroid tube coagu-lation test and the modified Lee-White test, using threetest tubes and an oiled syringe. Twenty of the controlgroup, all ambulatory psychiatric patients, were studiedwith the heparin tolerance test and the modified Lee-White test, using three saline-rinsed test tubes and asaline-rinsed syringe. The whole and dilute plasmaprothrombin times were studied in 28 ambulatory psy-chiatric patients. The Waugh-Ruddick test was studiedby means of 25 tests on a series of 15 persons in thecontrol group.

PROCEDURE

1. Modified Lee-White test

This test was run essentially as described by Lee andWhite (15), with the exception that three test tubes wereused instead of one. The second tube was not tilted untilclotting had occurred in the first, and the third tube wasnot tilted until clotting had occurred in the second. Thetime required for clotting in the third tube was taken asthe clotting time. This test was run in a constant tem-

671

NOBLE 0. FOWLER

perature water bath at 370 C as suggested by Quick (16).This test was also run on each patient using three drytest tubes and an oiled syringe.

2. Prothrombin time on whole and 12.5% plasma

The method described by Quick (17) was employed,using Squibb thromboplastin.

3. Lusteroid tube coagulation test

The method used was similar to that of Kadish (1, 2),except that three tubes were used instead of one, thereading being taken from the third tube. The tests were

run in a constant temperature water bath at 370 C. Bloodwas used only if obtained at the first attempt at venipunc-ture. Specimens were discarded if bubbling or frothingoccurred in the syringe. The lusteroid tubes were 12.5mm. in diameter. They were cleaned with sodium laurylsulfonate, as recommended by Kadish (1, 2).

4. Waugh-Ruddick heparin retarded coagulation test

This test often shows acceleration of coagulation whenthe Lee-White test is normal. The method employed was

that described by the original authors (8, 9), except thatall tests were run in a water bath at 22.5° C, as recom-

mended by Whittaker (18). This author found that a

rise in temperature caused much more rapid clotting inthe Waugh-Ruddick test; we have had a similar experi-ence. Whittaker selected 22.5° as the average winterroom temperature. Waugh and Ruddick did their testsat room temperature (8, 9). He found clotting timesin normals comparable to those of Waugh and Ruddickwhen the tubes containing the blood were placed in a

water bath at a temperature of 22.50 C (18). The cri-terion of increased coagulability in this test was thatsuggested by Ogura (19), namely, a coagulation timeof 60 minutes or less in each of the last three test tubesemployed.

5. Heparin tolerance test

Hagedorn and Barker's (11) modification of the testof deTakats (10) was employed. However, in consid-eration of the marked effect of temperature variationupon the clotting time of blood (16, 18), the tests were

done in a constant temperature water bath at 370 C.Three test tubes, each 8 mm. by 75 mm., were employed.

As suggested by Hagedorn and Barker (11), one sampleof blood was drawn 10 minutes after the injection of 25mg. of heparin in 2.5 cc. of solution intravenously. Theheparin was injected at a constant rate, requiring 30seconds for its administration.

RESULTS

The findings in the control group will be givenfirst. These are shown in Table I. In 50 con-trols, the range for the modified Lee-White test,using an oiled syringe, was from 9 to 23.5 minutes;the range for clotting in lusteroid was from 19 to40 minutes in this group. In 20 controls, the range

for the modified Lee-White test, using saline-rinsed glassware, was from 9 to 21 minutes. Themean for the 20 determinations was 11.58 min-utes when saline-rinsed tubes and syringes were

employed; the mean for 50 similar tests using an

oiled syringe and dry test tubes was 14.77 minutes.The standard deviation for the 50 determinationsusing oiled syringes was 3.34 minutes. The stand-ard deviation for the 20 determinations using a

saline-rinsed syringe was 4.36 minutes. The stand-ard error of the difference between the two means

of 14.77 minutes and 11.58 minutes was foundto be 1.08 minutes. The observed difference was

3.19 minutes. Since the observed difference inthe means was three times the standard error ofthe difference, this was felt to be a significantfinding. This difference would occur by chanceonly once in 370 tests. Ten minutes after the in-jection of 25 mg. of heparin intravenously, themodified Lee-White clotting time was increasedfrom two to slightly more than four times that be-fore heparin.

Among 28 control ambulatory psychiatric pa-

tients, the range for the prothrombin time run on

whole plasma was from 11.5 to 16.00 seconds. Therange for the prothrombin time on 12.5%o plasmawas 26 to 37 seconds; the range for the difference

TABLE I

Results of various coagulation tests in the control group

Nameof test Number of Range Meanpersons

Lee-White (oiled syringe) 50 9 to 23.5 min. 14.77 min.Lee-White (saline-rinsed syringe) 20 9 to 21 min. 11.58 min.Heparin tolerance test 20 Clotting time increased

2 to 4 timesLusteroid tube clotting test 50 19 to 40 min.Prothrombin time (whole plasma) 28 11.5 to 16 sec. 13.5 sec.Prothrombin time (12.5% plasma). 28 26 to 37 sec.

672

A STUDY OF CERTAIN ASPECTS OF BLOOD COAGULATION

TABLE II

Summary of findings in 29 postoperative cases

Operation p. 0. day Pro. time Pro. time Lee-White Lust. 370Ca R| WauddiOperation P. 0. old12.5% saline HPrn to RWuddh-whoe 15% 30C 370 C 3C 22.50 C

1. Append. 6 0 . 0 0 0 0 acc2. Hysterect. 6 0 0 0 0 0 acc3. Nephrect. 6 0 0 0 acc 0 acc4. Hernia 5 0 0 acc 0 0 05. Hernia 4 0 0 0 acc 0 acc6. Nephrect. 1 0 0 0 0 acc acc7. Hysterect. 3 0 acc acc 0 acc8. Append. 4 0 0 0 acc acc acc9. Hernia 4 acc 0 0 0 0 acc

10. Hernia 5 acc 0 0 acc 0 acc11. Gunshot colon 3 acc acc 0 0 0 acc12. Gunshot colon 4 0 0 0 0 0 acc13. Hernia 2 0 0 0 0 acc 014. Stab of abdomen 3 0 0 0 0 0 acc15. Append. 2 0 0 0 0 0 016. Hysterect. 9 0 0 0 acc acc17. Hysterect. 3 0 0 0 0 0 acc18. Append. 4 0 0 0 0 0 acc19. Hernia 3 0 acc acc acc 0 acc20. Hernia 4 0 0 0 0 0 021. Gunshot bowel 6 0 0 0 0 0 022. Gunshot bowel 2 0 0 0 0 0 acc23. Stab of bowel 6 0 0 0 acc acc 024. Stab of abdomen 2 0 0 0 acc 0 025. Append. 1 0 0 0 0 acc 026. Append. 1 0 0 0 acc 0 acc27. Gunshot bowel 4 acc 0 0 acc 0 028. Hernia 2 0 0 0 0 acc 029. Perf. ulcer 6 acc 0 0 0 0 acc

Note: acc indicates acceleration0 indicates no acceleration- indicates that the test was not run

between the whole and dilute plasma prothrombintimes was from 14 to 22 seconds.

Acceleration was considered present in any testwhen the result obtained was lower (i.e., less time)than was found in any control test of the samemethod. Of the 29 postoperative patients, accel-eration was found as follows: by Waugh-Ruddick,19 showed acceleration; by lusteroid, 10; by he-parin tolerance, six; by whole plasma prothrombintime, five; by 12.5%o plasma prothrombin time,three; by modified Lee-White test run in saline-rinsed glass, three. The results are summarizedin Table II. As seen in Table II, the correla-tion among the various tests was poor. Of 10 pa-tients showing abnormal tests in lusteroid, sevenshowed abnormal Waugh-Ruddick tests. How-ever, of six patients with abnormal heparin toler-ance, only two had accelerated Waugh-Ruddicktests, and only. two had accelerated- clotting inlusteroid. Of five patients having abnormally shortwhole plasma prothrombin times, four showed ab-normal Waugh-Ruddick tests, but only two had

accelerated clotting in lusteroid. The only pa-tient among the 29 tested who developed venousthrombosis had normal values for all tests. Thispatient was a 42 year old colored man, who had hada hernia repaired eight days prior to the thrombo-sis. The clotting tests were performed five daysprior to the occurrence of the thrombosis.

In order to ascertain how long the state of ac-

celerated coagulation may persist after operation,the same series of six tests was repeated on seven

of the 29 patients later in their postoperativecourses. The results are shown in Table III. Asmay be seen in the table, evidence of hypercoagu-lability was found as late as 12 days after operationin some instances.

Our clotting times after 25 mg. of heparin wereshorter than those obtained by Hagedorn andBarker (11), probably because our tests were

performed at 37° C, whereas their tests were per-formed at room temperature, which is generallymuch lower. In order to study the effect of tem-perature upon the results obtained in this test, we

673

NOBLE 0. FOWLER

TABLE III

Persistence of hypercoagulability following operation

Operation P. 0. day Prwhole Pro%time Lee-White Lust. Heparin tol. Ruddick

sec. sec. min. min. min. min.1. Hernia 2 15 30 91 311 171* 66

12 12.5 28.5 17i 26 36 72

2. Hysterect. 3 12.5 30.5 9 27 22 52*11 11.5 23.5* 13 29 42 48*

3. Gunshot bowel 4 10.5* 27 15 16* 49 6412 14 33.5 151 24 22* 64

4. Hernia 2 12.5 31.5 27 32 45 829 12.5 31 10 37 36 68

5. Perf. ulcer 6 11* 26 10 19 27 50*12 16.5 45 16 27 24* 58*

6. Append. 4 14 35 11 30 28 54*11 12.5 35 14 29 31 66

7. Gunshot liver 2 12 30 12 25 48 60*10 16 34.5 10 26 62t 58*

* Accelerated valuet Abnormal hypocoagulability

ran clotting times on samples of blood drawn 10minutes after giving 25. mg. of heparin intra-venously to each of 10 patients. One part of thesample was tested at 22.50, the average winterroom temperature, and the other portion was doneat 370. As may be seen in Table IV, the clottingtime was appreciably longer at 22.50 in all instancessave two. The average clotting time at 22.50 was26.5 minutes, and at 370, 13.4 minutes.

In order to study the effect of test tube sizeupon the results obtained in doing the Lee-Whiteclotting time, this test was done in glass tubes of

TABLE IV

The effect of temperature on theheparin tolerance test *

10 min. 10 min.after after

Before heparin heparinheparin 370 C 22.5° C

min. min. min.1. 71 22 382.9 11 523. 5 124. 7 12 27*5. 9 17 426. 6 18 187. 7 16 2018.- 10 149. 121 12

10.- 11 29* Blood specimens were drawn 10 minutes after 25 mg.

heparin intravenously, and were divided into two portions.Most of these are abnormal cases, i.e., postoperativecases.

two different sizes, one 8 mm. by 75 mm. and theother, 13 mm. by 100 mm., using identical samplesof blood. As shown in Table V, this procedurewas performed in 14 instances. It may be seenthat the clotting times were longer in the smalltubes in nine instances; the clotting times wereequal in one case; in the four remaining instancesthe clotting times were longer in the large tubes.Both Kadish (1) and Quick (16) found clottingtimes to be longer in large than in small glass tubes.

Whether or not the veins selected in performingthe heparin tolerance test affected the result wasalso studied. In 14 instances, the one tube Lee-

TABLE V

The effect of tube size on the three tubeLee-White clotting test

Tube 8 mm. by Tube 13 mm. by75 mm. 100 mm.

min. min.

1 12 102 13 103 15 134 11 95 11 116 12 617 16 128 13 1019 10 111

10 21 1111 10 1112 12 13113 10 10114 11 10o

674

A STUDY OF CERTAIN ASPECTS OF BLOODCOAGULATION

White test was performed 10 minutes after in-jecting 25 mg. of heparin intravenously. Bloodwas drawn almost simultaneously from the veininto which the heparin was given, and from thevein of the opposite arm. As seen in Table VI, theresults were the same or nearly so in seven in-stances. In six cases, the clotting times wereconsiderably longer on blood drawn from the veinwhich had received the heparin. In one instance,the blood drawn from the vein receiving the heparindid not clot after 12 hours. In one case, the clot-ting time was significantly shorter on blood takenfrom the vein receiving the heparin. This find-ing would seem to imply that in some instancesthere is retained in the vein injected with a foreignsubstance a larger proportion of substance than isfound in the general circulation.

TABLE VI

The effect of veins used upon the resultsof the heparin tolerance test

10 min. after 10 min. afterBefore heparin heparin same heparin op-

vein posite arm

min. min.1 6 391 min. 122 61 12 hours plus 103 81 14 min. 144 7 8* 9*5 7 731 50t6 7 13* 127 51 22j 248 5 12 139 5 21 12

10 _ 7* 5*11 7 10* 16*12 13 1013 20 12114 _ 10 11

* Acceleration.t 50 mg. heparin given instead of 25 mg.Note: Each patient received 25 mg. heparin intraven-

ously. After 10 minutes, one tube Lee-White clottingtimes were run on blood drawn almost simultaneouslyfrom the vein receiving the heparin and from the vein ofthe opposite arm.

DISCUSSION

The clotting mechanism in postoperative pa-tients was studied by Potts and Pearl (20), Daw-barn, Earlham, and Evans (21), and by ShapiroSherwin, and Gordimer (22). These investiga-tors found that the platelet count tended to fallduring the first three days postoperatively, but thatit began to rise on the sixth postoperative day, andreached a peak on the tenth to 14th postoperative

day. They found little change in the clotting timeas determined by the usual methods.

The abnormalities in the clotting mechanismfound by us do not parallel the changes in theplatelet count found by the above investigators; ingeneral we found abnormalities in clotting earlierin the postoperative course in this study. Notonly were the manifestations of derangements inthe clotting mechanism found earlier here, but thechange seemed to be greater earlier in the post-operative course than in the 10 to 14 day post-operative period when the maximum changes inthe platelet count were found. However, as seenin Table III some cases still showed accelerationof clotting as long as 12 days after operation.

The finding of accelerated coagulation as judgedby the Waugh-Ruddick test in 19 of the 29 post-operative patients is in keeping with the findingsof Waugh and Ruddick (9), whose study of post-operative patients revealed accelerated coagula-tion 24 hours after operation; this accelerationwas present seven days following operation. Ourfinding of acceleration of coagulation in two-thirdsof 29 postoperative patients, a finding which per-sisted at least 12 days after operation in somecases, may be considered in regards to possibleroutine postoperative anticoagulant therapy.

The Waugh-Ruddick test seemed most sensitiveof the tests employed in detecting the effect ofoperative trauma. In this connection, it is of in-terest to note that Flinn (23), following the ad-ministration of digitalis in man, found accelerationof coagulation as shown by the Waugh-Ruddicktest, whereas the Lee-White test and the pro-thrombin time on whole and dilute plasma showedno change. The extreme sensitivity of the Waugh-Ruddick test is shown by its acceleration in re-sponse to such procedures as operative trauma anddigitalization; this would seem to detract from itsvalue in the prediction or diagnosis of venousthrombosis. The non-specificity of this test isfurther illustrated by the fact that Waugh andRuddick found that bed rest alone caused accelera-tion of coagulation (9).

It is of interest to note the lack of correlationbetween the Waugh-Ruddick test and the heparintolerance test, since heparin is used to prolong thecoagulation of blood in both instances, in the for-mer in vitro and in the latter in vivo. In partialexplanation of this finding, the statements of

675

NOBLE 0. FOWLER

Best and Jaques (24) may be cited. These writersstate that if there is mixture of the same amountsof heparin and blood in vivo and in vitro, theclotting time will be longer in vivo if moderateamounts of heparin are used. Best and Jaquesalso state that the renal threshold for heparin isquite low, and that there is rapid excretion ofheparin into the urine if large doses are given in-travenously. Further, even.if heparin is not ex-

creted into the urine, it disappears rapidly from theblood. Thus, variation in the rate of heparindestruction and heparin excretion have to be reck-oned with in performing the heparin tolerancetest. Since these factors are not operative in theWaugh and Ruddick test, the failure of the twotests to correlate may be explained.

The lack of correlation among all the tests mightbe explained in part by the onset of anxiety duringthe period of the test. DeTakats (25) has men-

tioned adrenal stimulation, such as by anxiety or

fear, as a cause of increased coagulability of theblood. In a few cases we have found marked ac-

celeration of coagulation in lusteroid tubes andmarked increase of heparin tolerance when the pa-

tients were on the verge of an anxiety attack or

delirium tremens. Other factors, such as type ofanesthesia, the presence of an occult infection, or

the administration of barbiturates might explainthe variations in clotting from patient to patientin the postoperative group. However, since allsix tests were done within a space of two hours on

each postoperative patient, these last factors couldnot very well explain the lack of correlation in theindividual patient; i.e., these factors do not explainwhy some tests were normal and others abnormalin the identical patient at virtually the same time.

The prothrombin time of dilute plasma was lessoften abnormal than that of whole plasma. Thiswas especially notable in the three thrombophlebiticpatients, where the dilute plasma prothrombin timewas normal in every instance, yet the whole plasmaprothrombin time was shortened in all three. Thisfinding is similar to that of Mahoney and Sandrock(26), who found a greater increase in prothrombinactivity in their postoperative thrombophlebitis pa-

tients as shown by the test on whole plasma as com-

pared to dilute plasma. Hurn, Barker, and Mann(27) also found that the prothrombin time of diluteplasma gave no more information as to hypercoagu-lability than the test as run on whole plasma. Tuft

and Rosenfield (28) concluded that accelerated di-lute plasma prothrombin determinations could notbe used for prediction or diagnosis of thrombo-embolic disease. In contrast, Shapiro (13, 14, 22)and Brambel and Loker (12) found the diluteplasma prothrombin time more sensitive to thestate of hypercoagulability than the whole plasmaprothrombin time.

Mahoney and Sandrock (26) found in their 58normal postoperative patients a prolongation of theprothrombin time during the first three days fol-lowing surgery. They found acceleration duringthis period only in those patients who subsequentlydeveloped thrombophlebitis. In our group, of 13patients studied during the first three days post-operatively, none showed the prolongation of thewhole plasma prothrombin time described by Ma-honey and Sandrock (26). Further, the meanwhole plasma prothrombin time in these 13 post-operative patients was 12.8 seconds, as comparedwith a mean of 13.5 seconds for the 28 controls.Among the 28 controls, the standard deviation forthe whole plasma prothrombin time was 0.85 sec-onds. Among the 13 postoperative patients, thestandard deviation for the whole plasma pro-thrombin time was 1.45 seconds. The standarderror of the difference of the two means of 12.8and 13.5 seconds was found to be 0.471 seconds.The observed difference was 0.70 seconds. This isnot a significant difference.

Only one patient of this group of 29 postopera-tive patients developed thrombophlebitis after op-eration. The demonstration of frequent abnor-malities in the six tests in so many of the 28patients who did not subsequently develop throm-bophlebitis would seem to indicate that none ofthese tests could be used to predict which patient isgoing to develop venous thrombosis following anoperation. This statement is given weight by thefact that the only patient who did develop clinicalthrombophlebitis showed no abnormality in any ofthe six tests which had been run five days pre-viously.

It is thought worthy of comment that only one ofthe 27 patients having congestive heart failureshowed an acceleration of coagulation in lusteroidtubes, and only two as measured by the three tubeLee-White test. This finding may be comparedwith that of Cotlove and Vorzimer (29) who foundno shortening of the whole or dilute plasma pro-

676

A STUDY OF CERTAIN ASPECTS OF BLOOD COAGULATION

thrombin time in 20 patients with congestive heartfailure. It is also of interest to note that 25 ofour 27 patients were receiving digitalis therapyin maintenance doses, yet little evidence of accel-eration of clotting was found. Digitalis has beenfrequently mentioned as an accelerator of clotting(7, 25, 30). Ogura, Fetter, Blankenhorn, andGlueck (19) found some evidence of accelerationof coagulation in their patients as they were beingdigitalized, but found that this acceleration laterdisappeared in the majority, even though main-tenance digitalis therapy was continued. Cotloveand Vorzimer (29) found that the administrationof digitalis did not shorten the prothrombin timeas determined on whole plasma or dilute plasma.Moses (31) found no change in heparin toleranceafter the exhibition of digitalis. Flinn (23) foundthe Waugh-Ruddick test to show acceleration ofcoagulation after the administration of digitalis,but no acceleration was shown by the Lee-Whitetest or by the prothrombin time on whole or diluteplasma.

SUMMARY

Twenty-nine postoperative patients were stud-ied for evidence of accelerated coagulation. Nine-teen showed acceleration by the Waugh-Ruddicktest; ten showed acceleration in lusteroid tubes; sixshowed increased heparin tolerance; five showedshortening of the whole plasma prothrombin time;three showed shortening of the 12.5% plasma pro-thrombin time; three showed an acceleration by themodified Lee-White test.

The same tests were repeated on seven of thesepatients later in their postoperative courses andsome showed persistence of hypercoagulability onthe 12th postoperative day. The different tests didnot correlate. The only patient who developedsubsequent thrombophlebitis was normal to alltests. Three other patients with thrombophlebitisshowed acceleration of coagulation in lusteroidtubes and shortening of the whole plasma pro-thrombin time. They were normal with regardto dilute plasma prothrombin time and the Waugh-Ruddick test. Of 27 patients with congestive heartfailure, 25 of whom were receiving digitalis, onlyone showed acceleration of coagulation in lusteroidtubes, and two showed acceleration in the Lee-White test. Fourteen patients were given 25 mg.

of heparin intravenously. Ten minutes later, bloodsamples were drawn almost simultaneously fromthe arm vein into which the heparin had been in-jected, and from a vein in the opposite arm. Insix of these 14 patients, the clotting time was sig-nificantly longer in the blood sample taken fromthe vein receiving the heparin when compared tothe blood obtained from the opposite arm.

CONCLUSIONS

1. Hypercoagulability of the blood is a fre-quent finding in the postoperative state, and maypersist as long as 12 days after operation.

2. The correlation among the Waugh-Ruddicktest, prothrombin time on whole plasma, prothrom-bin time on 12.5%o plasma, lusteroid tube clottingtime, and heparin tolerance test is poor, suggestingthat they are concerned with different factors in-volved in the clotting process.

3. We could not predict which postoperativepatients were going to develop venous thrombosis.

4. When performing the heparin tolerance testand similar tests in which a substance is injectedintravenously, and a sample of blood is later drawnfor analysis, it is advisable to draw the blood fromthe vein of the opposite arm. The vein into whichthe substance is injected probably retains a largerproportion of that substance than is found in thegeneral circulation.

5. Of the tests studied, the Waugh-Ruddickheparin retarded coagulation test appears the mostsensitive. Its lack of specificity, however, wouldseem to indicate that it is of no value in predictingor diagnosing thrombophlebitis.

ACKNOWLEDGMENT

The writer wishes to acknowledge the very kindassistance of Dr. Helen Glueck.

BIBLIOGRAPHY

1. Kadish, A. H., Coagulation of the blood in lusteroidtubes; a study of normal persons and patients witharterial or venous thrombosis. Am. Heart J.,1947, 34, 212.

2. Kadish, A. H., Coagulation time of the blood inlusteroid tubes: a study of patients receiving di-cumarol. Am. Heart J., 1947, 34, 225.

3. Lozner, E. L., Taylor, F. H. L., and MacDonald, H.,The effect of foreign surfaces on blood coagulation.J. Clin. Invest., 1942, 21, 241.

6177

NOBLE 0. FOWLER

4. Tocantins, L. M., Influence of contacting surface oncoagulability and anticephalin activity of normaland hemophilic plasmas. Am. J. Physiol., 1945,143, 67.

S. Jaques, L. B., Fidlar, E., Feldsted, E. T., and Mac-Donald, A. G., Silicones and blood coagulation.Canad. M. A. J., 1946, 55, 26.

6. Hirschboeck, J. S., The effect of operation and ill-ness on clot retraction: description of a newmethod. J. Lab. & Clin. Med., 1948, 33, 347.

7. Massie, E., Stillerman, H. S., Wright, C. S., andMinnich, V., Effect of administration of digitalison coagulability of human blood. Arch. Int. Med.,1944, 74, 172.

8. Waugh, T. R., and Ruddick, D. W., A test for in-creased coagulability of the blood. Canad. M. A.J., 1944, 50, 547.

9. Waugh, T. R., and Ruddick, D. W., Studies on in-creased coagulability of the blood. Canad. M. A.J., 1944, 51, 11.

10. deTakats, G., and Gilbert, N. C., The response toheparin: a test of the clotting mechanism. J. A.M. A., 1943, 121, 1246.

11. Hagedorn, A. B., and Barker, N. W., Response ofpersons with and without intravascular thrombosisto a heparin tolerance test. Am. Heart J., 1948,35, 603.

12. Brambel, C. E., and Loker, F. F., Significance ofvariations of prothrombin activity of dilute plasma.Proc. Soc. Exper. Biol. & Med., 1943, 53, 218.

13. Shapiro, S., Hyperprothrombinemia, a premonitorysign of thromboembolization (description of amethod). Exper. Med. & Surg., 1944, 2, 103.

14. Shapiro, S., Sherwin, B., Redish, M., and Campbell,H. A., Prothrombin estimation; procedure andclinical interpretations. Proc. Soc. Exper. Biol. &Med., 1942, 50, 85.

15. Lee, R. I., and White, P. D., A clinical study of thecoagulation time of the blood. Am. J. M. Sc.,1913, 145, 495.

16. Quick, A. J., Rene, H. C., and Stefanini, M., Thevalue and limitations of the coagulation time inthe study of the hemorrhagic diseases. Blood,1948, 3, 1120.

17. Quick, A. J., The nature of bleeding in jaundice.J. A. M. A., 1938, 110, 1658.

18. Whittaker, J., The effect of temperature on the

Waugh and Ruddick test for increased coagulabil-ity of blood. Canad. M. A. J., 1945, 52, 185.

19. Ogura, J. H., Fetter, N. R., Blankenhorn, M. A.,and Glueck, H. I., Changes in blood coagulationfollowing coronary thrombosis measured by theheparin retarded clotting test (Waugh and Rud-dick test). J. Clin. Invest., 1946, 25, 586.

20. Potts, W. J., and Pearl, E., Study of platelet countand coagulation time of plasma and whole bloodfollowing operation. Surg., Gynec., & Obst., 1941,73, 492.

21. Dawbarn, R. Y., Earlham, F., and Evans, W. H.,The relation of the blood platelets to thrombosisafter operation and parturition. J. Path. & Bact.,1928, 31, 833.

22. Shapiro, S., Sherwin, B., and Gordimer, H., Post-operative thrombo-embolization. Ann. Surg., 1942,116, 175.

23. Flinn, J. H., The effect of digitalis on the coagulabil-ity of blood. Proc. Cent. Soc. Clin. Research,1948, 21, 100.

24. Best, C. H., and Jaques, L. B., Heparin in bloodclotting and thrombosis. Ann. N. Y. Acad. Sci-ences, 1948, 49, 501.

25. deTakats, G., Thromboembolism. Minn. Med., 1945,28, 843.

26. Mahoney, E. B., and Sandrock, R. S., The earlyrecognition of post-operative venous thrombosis.Bull. N. Y. Acad. Med., 1948, 24, 636.

27. Hurn, M., Barker, N. W., and Mann, F. D., Varia-tions in prothrombin and antithrombin in patientswith thrombosing tendencies. Am. J. Clin. Path.,1947, 17, 709.

28. Tuft, H. S., and Rosenfield, R. E., Significance ofthe accelerated reaction in determination of pro-thrombin time of diluted plasma. Am. J. Clin.Path., 1947, 17, 704.

29. Cotlove, E., and Vorzimer, J. J., Serial prothrombinestimations in cardiac patients: diagnostic andtherapeutic implications; use of dicumarol. Ann.Int. Med., 1946, 24, 648.

30. deTakats, G., Trump, R. A., and Gilbert, N. C., Theeffect of digitalis on the clotting mechanism.J. A. M. A., 1944, 125, 840.

31. Moses, C., The effect of digitalis, epinephrine, andsurgery on the response to heparin. J. Lab. &Clin. Med., 1945, 30, 603.

678